PISA is a unique global survey that examines how well prepared 15 year old students are for life after they have completed school. Approximately a half million students and 65 economies participated in this assessment that deals with reading, math, and science literacy.
The full 2012 PISA dataset can be found here.
The PISA data dictionary, which gives a brief explaination of each feature, can be found here.
For my analysis, I wanted to focus on how gender plays a role in not only student's mathematics scores but also student's attitude towards mathematics. I also wanted to take a look at if the parent's attitude towards children's mathematics differ among the gender.
The original data set contains 485,490 students and 636 features. Due the extremely large size of our features and the fact that we wanted to focus on gender, we have downsized our data set to include only 10 features.
My data cleaning process can be found here.
BRIEF DATA CLEANING PROCESS¶
For our data cleaning process, we focused on features that dealt with the country, gender, math scores, student's attitudes and student's perceived attitudes of parents towards mathematics. Student's attitudes can be broken down into categories: instrumental motivation, math anxiety, math self-concept, math interest, math work ethic, and math behavior. For each subcategory under student's attitudes, I assigned points and took the average to represent the overall score for each subcategory. For example, instumental motivation asks 4 questions (worthwhile for work, worthwhile for career changes, important for future study, and helps to get a job). Each of these questions can be answered by either 'Strongly Agree', 'Agree', 'Disagree','Strongly Disagree.' After focusing on only certain features as well as combining features, we are left with a total of 17 features.
OUR FINAL FEATURES
- country
- gender
- overall_score: student's overall math score
- The following are on a 4 point scale (4 points Strongly Agree, 3 points Agree, 2 points Disagree, 1 point Strongly Disagree):
- motivation: worthwhile for work, worthwhile for career changes, important for future study, helps to get a job
- anxiety: worry that it will be difficult, get very tense, get very nervous, feel helpless, worry about getting poor grades
- interest: enjoy reading, look forward to lessons, enjoy maths, interested
- work_ethic: homework completed on time, work hard on homework, prepared for exams, student hard for quizzes, study until I understand everything, pay attention in class, listen in classes, avoid distractions when studying, keep work organized
- parents: student's perceived view of parent's attitude towards mathematics; parents believe studying mathematics is important, parents believe mathematics is important for career, parents like mathematics
- The following are on a 4 point scale (4 points Almost of Almost Always, 3 points Often, 2 points Sometimes, 1 point Never or rarely):
- behavior: talk about maths with friends, help friends with mat, extracurricular activity,participates in competition, study more than 2 extra hours a day, play chess, computer programming, participates in math club
NOTE: Features under motivation, anxiety, interest, work_ethic, parents, and behavior all fall in a 4 point system scale. It is important to note that anxiety's scale shows 4 points being a negative behavior and 1 point shows a positive behavior (student's lack of anxiety towards mathematics). Contrary, the remainder of features here show the opposite. For example, interest's 4 points shows that a student is highly interested in math and 1 points show s student's lack of interest.
DATA EXPLORATION¶
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sb
%matplotlib inline
plt.style.use('fivethirtyeight')
df_pisa=pd.read_csv('pisa2012_clean.csv')
df_pisa.head()
#overview of data shape and composition
print(df_pisa.shape,'\n')
print(df_pisa.info())
df_pisa[['country','gender']]=df_pisa[['country','gender']].astype('category')
print(df_pisa.describe())
What is the structure of your dataset?
There are 485,490 students in the dataset with 17 features (country, gender, overall math score, motivation, anxiety, interest, work ethic, behavior, self, and parents). Country and Gender are of category type and the remainder of the features are numeric.
What is/are the main feature(s) of interest in your dataset?
For my analysis, I am interested in seeing the effects of gender has on students' mathematical skills. I not only wanted to take into account the students' math scores, but also the students and parent's feelings towards Mathematics.
From my own experience of studying Mathematics in college, I have seen a large gender gap in my higher math classes where the male student population far exceeds that of female students. This makes me wonder if the large gender gap is due to having a gap in scores and/or attitude between gender. In addition, I would also like to know if the Mathematic Skills and Attitudes between Gender across different countries as well if a student's perceived attitude of their parents differ between gender.
What features in the dataset do you think will help support your investigation into your feature(s) of interest?
I expect that there will be a higher overall test and subsection score among males compared to that of females. For test scores, we will be looking at the score_overall. I also expect there to be a higher attitude towards mathematics among males compared to that of the females. For attitudes, we will be looking at the following features: motivation, anxiety, interest, work_ethic, behavior, and self. All attitude features, with the exception of 'anxiety', show that the higher the score is, the better the student's attitude is towards Mathematics.
Univariate Exploration¶
SCORES
#Overall Score
binsize=50
bins=np.arange(0,df_pisa['score_overall'].max()+binsize,binsize)
plt.figure(figsize=[12,6])
plt.hist(data= df_pisa,x='score_overall',bins=bins)
plt.title('OVERALL MATH SCORES', size=25)
plt.xlabel('Math Score')
plt.ylabel('Number of Students');
plt.savefig('hist_score_overall.png')
The Overall Score looks pretty much normally distributed with a peak at around 500.
GENDER
plt.pie(df_pisa['gender'].value_counts(), labels=df_pisa['gender'].value_counts().index,startangle=90,counterclock=False)
plt.axis('square')
plt.title('STUDENTS BY GENDER');
print('Male Students: {:.4f}%'.format(df_pisa['gender'].value_counts()['Male']/df_pisa.shape[0]))
print('Female Students: {:.4f}%'.format(df_pisa['gender'].value_counts()['Female']/df_pisa.shape[0]))
Gender is almost evenly split.
COUNTRY
order= df_pisa['country'].value_counts().index
plt.figure(figsize=[20,30])
sb.countplot(data=df_pisa,y='country',order=order)
plt.title('NUMBER OF STUDENTS BY COUNTRY',size=20);
plt.figure(figsize=(20,8))
sb.boxplot(df_pisa['country'].value_counts(), color='pink')
sb.swarmplot(df_pisa['country'].value_counts(), color='crimson')
plt.title('NUMBER OF STUDENTS BY COUNTRY', size=30);
Most countries have between 5000ish to 7000ish students taking the survey. The boxplot also shows that there are a few outliers. For example, Mexico and Italy have over 30,000 students and Liechtenstein has well below 1000 students.
df_pisa['country'].value_counts().describe()
ATTITUDE
#Subsection Attitude
attitude=['motivation', 'anxiety','interest', 'work_ethic', 'behavior', 'self']
fig, ax= plt.subplots(nrows=2,ncols=3,figsize=[20,12])
#binsizes were determined by number of questions per section.
#binsizes = 1/(number of questions per section)
binsizes=[1/4,1/5,1/4,1/9,1/8,1/5]
ax=ax.flatten()
i=0
for feature in attitude:
bins=np.arange(min(df_pisa[feature]), max(df_pisa[feature]) +binsizes[i], binsizes[i])
ax[i].hist(data=df_pisa, x=feature, bins=bins)
ax[i].set_xlabel('Score')
ax[i].set_ylabel('Number of Students')
ax[i].set_title(feature)
i+=1
under the motivation feature, it looks like the graph is slighly skewed to the left with spikes at points 3-3.25 and 3.75-4.
anxiety is almost normally distributed
work_ethic is slightly skewed to the left with a spike at around 3 points.
behavior is skewed to the right showing that most students do not take extra actions towards Mathematics such as taking about math with friends, play chess, or computer programming
PARENTS
binsize=1/3
bins=np.arange(df_pisa['parents'].min(),df_pisa['parents'].max() +binsize, binsize)
plt.hist(data=df_pisa, x='parents', bins=bins)
plt.xlabel('score')
plt.ylabel('number of students')
plt.title('STUDENT\'S PERCEPTION OF PARENTAL ATTITUDE TOWARDS MATHEMATICS');
The graph is slightly skewed to the left showing that most students have a positive view of their parent's attitudes towards Mathematics
What if we now just take a look at just the top percentile of students and see how gender is distributed?¶
df_75=df_pisa.query('score_overall>=@df_pisa.score_overall.quantile(.75)')
df_90=df_pisa.query('score_overall>=@df_pisa.score_overall.quantile(.90)')
df_95=df_pisa.query('score_overall>=@df_pisa.score_overall.quantile(.95)')
df_99=df_pisa.query('score_overall>=@df_pisa.score_overall.quantile(.99)')
percent=[75,90,95,99]
fig, ax=plt.subplots(nrows=2, ncols=2, figsize=(10,10))
ax=ax.flatten()
for i in range(4):
plt.sca(ax[i])
df=eval('df_'+str(percent[i]))
plt.pie(df['gender'].value_counts(), labels=df['gender'].value_counts().index,startangle=90,counterclock=False,autopct='%.2f')
plt.axis('square')
plt.title('TOP '+ str(percent[i])+'th PERCENTILE BY GENDER')
This is actually very interesting to see. Notice how as the top nth percentile increases, the gender gap increases as well.
BIVARIATE EXPLORATION¶
Gender by Each Country
order= df_pisa['country'].value_counts().index
plt.figure(figsize=[20,40])
sb.countplot(data=df_pisa,y='country',order=order,hue='gender')
plt.title('Number of Students by Country',size=35);
While most countries show an almost 50/50 split between gender, there are some countries like Mexico, Thailand and Israel that show an very uneven split. Let's take a closer look at this!
country=df_pisa['country'].unique().tolist()
country.sort()
female=df_pisa.query('gender=="Female"').groupby('country').size()
male=df_pisa.query('gender=="Male"').groupby('country').size()
# we must make sure that the index matches for all
country==female.index.tolist()==male.index.tolist()
df_gender=pd.DataFrame({'country': country, 'female': female.values,'male':male.values})
df_gender['total_pop']=df_gender['female']+df_gender['male']
df_gender['female_percent']=100*(df_gender['female']/df_gender['total_pop'])
df_gender['male_percent']=100*(df_gender['male']/df_gender['total_pop'])
df_gender['diff_percent']=(np.absolute((df_gender['female_percent']-df_gender['male_percent'])))
df_gender.head(10)
df_gender['diff_percent'].describe()
sb.swarmplot(data=df_gender,x='diff_percent',color='crimson')
sb.boxplot(data=df_gender, x='diff_percent',color='pink')
plt.title('PERCENT DIFFERENCE IN GENDER BY COUNTRY')
plt.xlabel('percent');
Almost half of the countries have a difference in gender of about 2% and lower for students. It even looks like 2 countries have a difference of more than 10%. Why is there such a large gap in gender for students taking these exams for these countries?
df_gender.sort_values('diff_percent',ascending=False).head(10)
If we take a look at the top 10 countries with the most gender differences, 7 out of 10 countries are female dominated.
GENDER BY OVERALL SCORES
binsize=25
bins=np.arange(min(df_pisa['score_overall']),max(df_pisa['score_overall'])+binsize, binsize)
plt.title('OVERALL SCORE')
plt.hist(data=df_pisa.query('gender=="Female"'),x='score_overall', alpha=.4,bins=bins,label='Female')
plt.hist(data=df_pisa.query('gender=="Male"'),x='score_overall', alpha=.4,bins=bins, label='Male')
plt.legend();
The Males students consistently score better than the Female students.
GENDER BY STUDENT'S PERCEPTION OF PARENTAL ATTITUDE TOWARDS MATHEMATICS
sb.boxplot(data=df_pisa, y='parents', x='gender')
;
df_pisa.query('gender=="Female"').parents.describe(), df_pisa.query('gender=="Male"').parents.describe()
Note that 5000 less Male students did not answer all the questions under Parental View
It looks like the there is only a slight difference in male/female perception of their parent's towards mathematics. For both male and female, the 25%, 50%, and 75% are surprisenly the same.
plt.hist(df_pisa.query('gender=="Female"')['parents'],bins=np.arange(1,4+1/3,1/3), label='Female', alpha=.4)
plt.hist(df_pisa.query('gender=="Male"')['parents'],bins=np.arange(1,4+1/3,1/3), label='Male', alpha=.4)
plt.legend()
plt.title('PARENTAL VIEW OF MATHEMATICS BY GENDER')
plt.xlabel('score')
plt.ylabel('number of students');
Since there is an uneven amount of female and males students who answered all questions under Parental View, let's go ahead and take a look at the density.
sb.kdeplot(data=df_pisa.query('gender=="Female"')['parents'], shade=True, color='dodgerblue', bw=1/4, label='Female')
sb.kdeplot(data=df_pisa.query('gender=="Male"')['parents'], shade=True, color='pink', bw=1/4, label='Male')
plt.title('PARENTAL VIEW OF MATHEMATICS BY GENDER (DENSITY CURVE)')
plt.xlabel('score')
plt.ylabel('number of students');
Again we see that only a slight amount of male students view their parent's having a higher perception of mathematics.
GENDER BY ATTITUDE
#Subsection Attitude
attitude=['motivation', 'anxiety','interest', 'work_ethic', 'behavior', 'self']
fig, ax= plt.subplots(nrows=2,ncols=3,figsize=[20,12])
#binsizes were determined by number of questions per section.
#binsizes = 1/(number of questions per section)
binsizes=[1/4,1/5,1/4,1/9,1/8,1/5]
ax=ax.flatten()
i=0
for feature in attitude:
bins=np.arange(min(df_pisa[feature]), max(df_pisa[feature]) +binsizes[i], binsizes[i])
ax[i].hist(data=df_pisa.query('gender=="Female"'), x=feature, bins=bins,label='Female', alpha=.4)
ax[i].hist(data=df_pisa.query('gender=="Male"'),x=feature, bins=bins,label='Male', alpha=.4)
ax[i].set_xlabel('Score')
ax[i].set_ylabel('Number of Students')
ax[i].set_title(feature)
ax[i].legend()
i+=1
plt.savefig('attitudes_gender.png')
Males students have a high score under motivation, interest, and self. Female students seem to score higher under anxiety as well as work_ethic
OVERALL SCORES BY ATTITUDE
#attitude=['motivation', 'anxiety','interest', 'work_ethic', 'behavior', 'self']
fig, ax= plt.subplots(nrows=2,ncols=3,figsize=[20,12])
ax=ax.flatten()
i=0
for feature in attitude:
plt.sca(ax[i])
sb.scatterplot(x=df_pisa[feature],y=df_pisa['score_overall'],alpha=.05)
ax[i].set_ylabel('Score: {}'.format('Overall Score'))
ax[i].set_xlabel('Attitude Score')
ax[i].set_title('Overall Score by {}'.format(feature.title()))
i+=1
Overall score does not seem to be highly correlated with any features under attitude
It would be interesting later to see if there does exist a correlation between different 'bins' under Overall Score with each feature under Attitude.
sb.heatmap(data=df_pisa[['score_overall','motivation', 'anxiety','interest', 'work_ethic', 'behavior', 'self']].corr(),
center=0, cmap="RdBu_r",annot=True, vmin=-1, vmax=1);
We can also see that there is a moderate positive correlation between self and interest, motivation and interest, and a negative moderate correlation between self and anxiety.
OVERALL SCORE BY PARENTAL VIEW OF MATHEMATICS
sb.scatterplot(x=df_pisa['parents'],y=df_pisa['score_overall'],alpha=.05);
df_pisa[['parents','score_overall']].corr()
Again, there does not seem to have a correaltion between parents and score_overall
ATTITUDE BY PARENTAL VIEW
sb.heatmap(data=df_pisa[attitude+['parents']].corr(), center=0, cmap="RdBu_r",annot=True, vmin=-1, vmax=1);
There is a moderate positive corretion of how the student's perceive the parent's attitude towards mathematics and the student's motivation for mathematics.
MULTIVARIATE EXPLORATION¶
MOTIVATION SCORE BY GENDER BY COUNTRY
g=sb.FacetGrid(data=df_pisa, col='country', hue='gender', col_wrap=4)
g.map(sb.kdeplot, 'motivation', bw=1/4)
for ax, c in zip(g.axes.flat, country):
ax.axvline(x=df_pisa.query('country==@c').motivation.quantile(.5), color='grey')
ax.axvline(x=df_pisa.query('country==@c').motivation.quantile(.75), color='maroon', alpha=.75)
ax.legend()
Due to the fact that most countries do not have an even 50/50 split among gender, I wanted to depict the motivation scores by density.
The grey vertical lines shown above shows the 50 percentile motivation score for each country meanwhile the maroon lines depict the 75 percentile. Notice that almost all countries show the male students outweighing female students in motivation scores.
ANXIETY SCORE BY GENDER BY COUNTRY
g=sb.FacetGrid(data=df_pisa, col='country', hue='gender', col_wrap=4)
g.map(sb.kdeplot, 'anxiety', bw=1/5)
for ax, c in zip(g.axes.flat, country):
ax.axvline(x=df_pisa.query('country==@c').anxiety.quantile(.5), color='grey')
ax.axvline(x=df_pisa.query('country==@c').anxiety.quantile(.75), color='maroon', alpha=.75)
ax.legend()
Almost all of the countries show females have a higher than average or similar axienty score compared to their male counterpart.
INTEREST SCORE BY GENDER BY COUNTRY
g=sb.FacetGrid(data=df_pisa, col='country', hue='gender', col_wrap=4)
g.map(sb.kdeplot, 'interest', bw=1/4)
for ax, c in zip(g.axes.flat, country):
ax.axvline(x=df_pisa.query('country==@c').interest.quantile(.5), color='grey')
ax.axvline(x=df_pisa.query('country==@c').interest.quantile(.75), color='maroon', alpha=.75)
ax.legend()
All of the countries show that the male proportion scored the most compared to the females.
WORK ETHIC SCORE BY GENDER BY COUNTRY
g=sb.FacetGrid(data=df_pisa, col='country', hue='gender', col_wrap=4)
g.map(sb.kdeplot, 'work_ethic', bw=1/4)
for ax, c in zip(g.axes.flat, country):
ax.axvline(x=df_pisa.query('country==@c').work_ethic.quantile(.5), color='grey')
ax.axvline(x=df_pisa.query('country==@c').work_ethic.quantile(.75), color='maroon', alpha=.75)
ax.legend()
Under Work Ethic, the females seem to score higher (or about the same) in proportion comparted to the males.
BEHAVIOR SCORE BY GENDER BY COUNTRY
g=sb.FacetGrid(data=df_pisa, col='country', hue='gender', col_wrap=4)
g.map(sb.kdeplot, 'behavior', bw=.125)
for ax, c in zip(g.axes.flat, country):
ax.axvline(x=df_pisa.query('country==@c').behavior.quantile(.5), color='grey')
ax.axvline(x=df_pisa.query('country==@c').behavior.quantile(.75), color='maroon', alpha=.75)
ax.legend()
Under behavior, males seem to have a higher score. A majority of these countries also have scores skewed the the right.
SELF SCORE BY GENDER BY COUNTRY
g=sb.FacetGrid(data=df_pisa, col='country', hue='gender', col_wrap=4)
g.map(sb.kdeplot, 'self', bw=1/5)
for ax, c in zip(g.axes.flat, country):
ax.axvline(x=df_pisa.query('country==@c').self.quantile(.5), color='grey')
ax.axvline(x=df_pisa.query('country==@c').self.quantile(.75), color='maroon', alpha=.75)
ax.legend()
PARENTAL SCORE BY GENDER BY COUNTRY
g=sb.FacetGrid(data=df_pisa, col='country', hue='gender', col_wrap=4)
g.map(sb.kdeplot, 'parents', bw=1/3)
for ax, c in zip(g.axes.flat, country):
ax.axvline(x=df_pisa.query('country==@c').parents.quantile(.5), color='grey')
ax.axvline(x=df_pisa.query('country==@c').parents.quantile(.75), color='maroon', alpha=.75)
ax.legend()
LET'S GO AHEAD AND TAKE A LOOK AT THE CHANGES IN OUR DATA AS THE TOP PERCENTILE UNDER OVERALL SCORE INCREASES
df=pd.DataFrame(columns=['nth_percentile','m_prop','f_prop',
'avg_motivation','avg_anxiety','avg_interest','avg_work_ethic','avg_parents','avg_behavior','avg_self',
'm_motivation','m_anxiety','m_interest','m_work_ethic','m_parents','m_behavior','m_self',
'f_motivation','f_anxiety','f_interest','f_work_ethic','f_parents','f_behavior','f_self'])
for i in range(100):
n=i*.01
s=df_pisa.score_overall.quantile(n)
df_top=df_pisa.query('score_overall>=@s')
m_prop= df_top.gender.value_counts()['Male']/df_top.gender.value_counts().sum()
f_prop= df_top.gender.value_counts()['Female']/df_top.gender.value_counts().sum()
motivation=df_top.motivation.mean()
anxiety=df_top.anxiety.mean()
interest=df_top.interest.mean()
work_ethic=df_top.work_ethic.mean()
parents=df_top.parents.mean()
behavior= df_top.behavior.mean()
self=df_top.self.mean()
df_f=df_top.query('gender=="Female"')
f_motivation=df_f.motivation.mean()
f_anxiety=df_f.anxiety.mean()
f_interest=df_f.interest.mean()
f_work_ethic=df_f.work_ethic.mean()
f_parents=df_f.parents.mean()
f_behavior= df_f.behavior.mean()
f_self=df_f.self.mean()
df_m=df_top.query('gender=="Male"')
m_motivation=df_m.motivation.mean()
m_anxiety=df_m.anxiety.mean()
m_interest=df_m.interest.mean()
m_work_ethic=df_m.work_ethic.mean()
m_parents=df_m.parents.mean()
m_behavior= df_m.behavior.mean()
m_self=df_m.self.mean()
df=df.append({'nth_percentile':n,'m_prop':m_prop,'f_prop':f_prop,
'avg_motivation':motivation,'avg_anxiety':anxiety,'avg_interest':interest,
'avg_work_ethic':work_ethic,'avg_parents':parents,'avg_behavior':behavior,'avg_self':self,
'm_motivation':m_motivation,'m_anxiety':m_anxiety,'m_interest':m_interest,
'm_work_ethic':m_work_ethic,'m_parents': m_parents,'m_behavior':m_behavior,'m_self':m_self,
'f_motivation':f_motivation,'f_anxiety':f_anxiety,'f_interest':f_interest,
'f_work_ethic':f_work_ethic,'f_parents':f_parents,'f_behavior':f_behavior,'f_self':f_behavior},
ignore_index=True)
df.head()
plt.figure(figsize=(10,8))
sb.lineplot(data=df,x='nth_percentile',y='f_prop', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_prop',label='Male')
plt.legend()
plt.title('GENDER PROPORTION FOR TOP PERCENTILE')
plt.ylabel('proportion')
plt.xlabel('nth percentile of overall math score');
We can see clearly here that as the top percentile increases, so does the gender gap. Initially, the female proportion was higher (due to the fact that there was a slightly more female students taking the survey as compared to the male students). But by the time we arrived to the top (approximately) 20th percentile, the amount of female students was equal to the amount of male students. All throughout the increase of top percentiles, we see that the proportion of male students strictly increases as well. Notice that the further we get along the top percentile of students, the gender gap's increase accelerates.
sb.lineplot(data=df,x='nth_percentile',y='f_motivation', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_motivation', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_motivation', label='Average Total')
plt.legend()
plt.title('AVERAGE MOTIVATION SCORE FOR TOP PERCENTILE')
plt.ylim(bottom=1)
plt.ylabel('motivation score')
plt.xlabel('nth percentile of overall math score');
The average motivation score for males is consistently higher that the average motivation score for females.
sb.lineplot(data=df,x='nth_percentile',y='f_anxiety', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_anxiety', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_anxiety', label='Average Total')
plt.legend()
plt.title('AVERAGE ANXIETY SCORE FOR TOP PERCENTILE')
plt.ylim(bottom=1)
plt.ylabel('anxiety score')
plt.xlabel('nth percentile of overall math score');
The average anxiety score for females is consistently higher than the average anxiety score for males. As the top percentile increases, the anxiety scores for both gender decreases.
sb.lineplot(data=df,x='nth_percentile',y='f_interest', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_interest', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_interest', label='Average Total')
plt.legend()
plt.title('AVERAGE INTEREST SCORE FOR TOP PERCENTILE')
plt.ylim(bottom=1)
plt.ylabel('interest score')
plt.xlabel('nth percentile of overall math score');
The average interest score for males is consistently higher that the average interest score for females.
plt.figure(figsize=(20,8))
plt.subplot(1,2,1)
sb.lineplot(data=df,x='nth_percentile',y='f_work_ethic', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_work_ethic', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_work_ethic', label='Average Total')
plt.legend()
plt.title('AVERAGE WORK ETHIC SCORE FOR TOP PERCENTILE')
plt.ylim(bottom=1)
plt.ylabel('work ethic score')
plt.xlabel('nth percentile of overall math score')
plt.subplot(1,2,2)
sb.lineplot(data=df,x='nth_percentile',y='f_work_ethic', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_work_ethic', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_work_ethic', label='Average Total')
plt.legend()
plt.title('CLOSER LOOK: AVERAGE WORK ETHIC SCORE FOR TOP PERCENTILE')
plt.ylabel('work ethic score')
plt.xlabel('nth percentile of overall math score');
The average work ethic score for females is consistently higher that the average work score for males.
plt.figure(figsize=(20,8))
plt.subplot(1,2,1)
sb.lineplot(data=df,x='nth_percentile',y='f_parents', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_parents', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_parents', label='Average Total')
plt.legend()
plt.title('AVERAGE PARENTS SCORE FOR TOP PERCENTILE')
plt.ylim(bottom=1)
plt.ylabel('parents score')
plt.xlabel('nth percentile of overall math score')
plt.subplot(1,2,2)
sb.lineplot(data=df,x='nth_percentile',y='f_parents', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_parents', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_parents', label='Average Total')
plt.legend()
plt.title('CLOSER LOOK: AVERAGE PARENTS SCORE FOR TOP PERCENTILE')
plt.ylabel('parent score')
plt.xlabel('nth percentile of overall math score');
The average parent score for males is consistently higher that the average motivation score for females. Notice how there is a sharp dip in parent score for males students at around the 85th percentile.
sb.lineplot(data=df,x='nth_percentile',y='f_behavior', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_behavior', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_behavior', label='Average Total')
plt.legend()
plt.title('AVERAGE BEHAVIOR SCORE FOR TOP PERCENTILE')
plt.ylim(bottom=1)
plt.ylabel('behavior score')
plt.xlabel('nth percentile of overall math score');
The average behavior score for males is consistently higher that the average behavior score for females.
sb.lineplot(data=df,x='nth_percentile',y='f_self', label='Female')
sb.lineplot(data=df,x='nth_percentile',y='m_self', label='Male')
sb.lineplot(data=df,x='nth_percentile',y='avg_self', label='Average Total')
plt.legend()
plt.title('AVERAGE SELF SCORE FOR TOP PERCENTILE')
plt.ylim(bottom=1)
plt.ylabel('self score')
plt.xlabel('nth percentile of overall math score');
The average self score for males is consistently higher that the average self score for females. Note how there is a large gap between how females feel about themselves regarding their math skills as compared to how males feel about themselves in their math skills.